Student develops CS2 player-tracking system to identify repeat cheaters

Christopher B. Didriksen of the Norwegian University of Science and Technology developed a system that could help identify repeat cheaters in Counter-Strike 2 by the way they control the game. He made the research the subject of his master’s thesis, which he completed with an A grade.

The system analyses separate mouse and keyboard signals in recordings of competitive matches, known as demos. Those patterns form a behavioural profile that, according to the research, can remain stable across maps, sessions and settings.

Keyboard profile reached 98% accuracy

In a dataset of more than 1,000 players, mouse-movement biometrics identified the correct player every time, although often by a narrow margin. Keyboard analysis identified the right player 98% of the time, while combining both signals allowed different accounts to be linked to the same person.

The profile remained stable across matches played months apart and after a sensitivity change from 800 to 640 eDPI. Didriksen says the system found previously unreported smurf accounts and is fast enough to check each new match against all of CS2’s monthly players.

Proposed alongside Valve’s systems

The researcher proposes using movement tracking alongside Valve Anti-Cheat and Trust Factor. The approach is intended to answer whether multiple accounts belong to the same person—an issue Didriksen says existing systems do not fully resolve.

The method is not foolproof. Shared accounts can distort the behavioural profile, and the system effectively removes the possibility that a banned cheater might return on a new account and play fairly.

Reddit users joined the research earlier this year, with some disclosing their own smurf accounts. Didriksen described that material as among the most valuable data collected for the thesis.